mirrors/ArchiveBox
1
0
Fork 0
mirror of https://github.com/ArchiveBox/ArchiveBox.git synced 2025-05-19 17:35:09 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions

migrate SnapshotTag to use new snapshot id

Browse source
This commit is contained in:
Nick Sweeting 2024-08-19 19:42:25 -07:00
parent 74c11d41be
commit 344e902fc6
No known key found for this signature in database
12 changed files with 251 additions and 6 deletions
Download patch file Download diff file Expand all files Collapse all files

2
archivebox/core/migrations/0044_alter_archiveresult_snapshot_alter_tag_uuid_and_more.py
View file

@ -22,7 +22,7 @@ class Migration(migrations.Migration):
name='SnapshotTag',
fields=[
('id', models.AutoField(primary_key=True, serialize=False)),
('snapshot', models.OneToOneField(on_delete=django.db.models.deletion.CASCADE, to='core.snapshot')),
('snapshot', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, to='core.snapshot')),
('tag', models.ForeignKey(on_delete=django.db.models.deletion.CASCADE, to='core.tag')),
],
options={

19
archivebox/core/migrations/0045_alter_snapshot_old_id.py Normal file
View file

@ -0,0 +1,19 @@
# Generated by Django 5.0.6 on 2024-08-20 01:54
import uuid
from django.db import migrations, models
class Migration(migrations.Migration):
dependencies = [
('core', '0044_alter_archiveresult_snapshot_alter_tag_uuid_and_more'),
]
operations = [
migrations.AlterField(
model_name='snapshot',
name='old_id',
field=models.UUIDField(default=uuid.uuid4, editable=False, primary_key=True, serialize=False, unique=True),
),
]

30
archivebox/core/migrations/0046_alter_archiveresult_snapshot_alter_snapshot_id_and_more.py Normal file
View file

@ -0,0 +1,30 @@
# Generated by Django 5.0.6 on 2024-08-20 01:55
import django.db.models.deletion
import uuid
from django.db import migrations, models
class Migration(migrations.Migration):
dependencies = [
('core', '0045_alter_snapshot_old_id'),
]
operations = [
migrations.AlterField(
model_name='archiveresult',
name='snapshot',
field=models.ForeignKey(db_column='snapshot_id', on_delete=django.db.models.deletion.CASCADE, to='core.snapshot', to_field='id'),
),
migrations.AlterField(
model_name='snapshot',
name='id',
field=models.UUIDField(default=uuid.uuid4, primary_key=True, serialize=False, unique=True),
),
migrations.AlterField(
model_name='snapshot',
name='old_id',
field=models.UUIDField(default=uuid.uuid4, editable=False, unique=True),
),
]

24
archivebox/core/migrations/0047_alter_snapshottag_unique_together_and_more.py Normal file
View file

@ -0,0 +1,24 @@
# Generated by Django 5.0.6 on 2024-08-20 02:16
import django.db.models.deletion
from django.db import migrations, models
class Migration(migrations.Migration):
dependencies = [
('core', '0046_alter_archiveresult_snapshot_alter_snapshot_id_and_more'),
]
operations = [
migrations.AlterField(
model_name='archiveresult',
name='snapshot',
field=models.ForeignKey(db_column='snapshot_id', on_delete=django.db.models.deletion.CASCADE, to='core.snapshot', to_field='id'),
),
migrations.AlterField(
model_name='snapshottag',
name='tag',
field=models.ForeignKey(db_column='tag_id', on_delete=django.db.models.deletion.CASCADE, to='core.tag'),
),
]

24
archivebox/core/migrations/0048_alter_archiveresult_snapshot_and_more.py Normal file
View file

@ -0,0 +1,24 @@
# Generated by Django 5.0.6 on 2024-08-20 02:17
import django.db.models.deletion
from django.db import migrations, models
class Migration(migrations.Migration):
dependencies = [
('core', '0047_alter_snapshottag_unique_together_and_more'),
]
operations = [
migrations.AlterField(
model_name='archiveresult',
name='snapshot',
field=models.ForeignKey(db_column='snapshot_id', on_delete=django.db.models.deletion.CASCADE, to='core.snapshot'),
),
migrations.AlterField(
model_name='snapshottag',
name='snapshot',
field=models.ForeignKey(db_column='snapshot_id', on_delete=django.db.models.deletion.CASCADE, to='core.snapshot', to_field='old_id'),
),
]

22
archivebox/core/migrations/0049_rename_snapshot_snapshottag_snapshot_old_and_more.py Normal file
View file

@ -0,0 +1,22 @@
# Generated by Django 5.0.6 on 2024-08-20 02:26
from django.db import migrations
class Migration(migrations.Migration):
dependencies = [
('core', '0048_alter_archiveresult_snapshot_and_more'),
]
operations = [
migrations.RenameField(
model_name='snapshottag',
old_name='snapshot',
new_name='snapshot_old',
),
migrations.AlterUniqueTogether(
name='snapshottag',
unique_together={('snapshot_old', 'tag')},
),
]

19
archivebox/core/migrations/0050_alter_snapshottag_snapshot_old.py Normal file
View file

@ -0,0 +1,19 @@
# Generated by Django 5.0.6 on 2024-08-20 02:30
import django.db.models.deletion
from django.db import migrations, models
class Migration(migrations.Migration):
dependencies = [
('core', '0049_rename_snapshot_snapshottag_snapshot_old_and_more'),
]
operations = [
migrations.AlterField(
model_name='snapshottag',
name='snapshot_old',
field=models.ForeignKey(db_column='snapshot_old_id', on_delete=django.db.models.deletion.CASCADE, to='core.snapshot', to_field='old_id'),
),
]

40
archivebox/core/migrations/0051_snapshottag_snapshot_alter_snapshottag_snapshot_old.py Normal file
View file

@ -0,0 +1,40 @@
# Generated by Django 5.0.6 on 2024-08-20 02:31
import django.db.models.deletion
from django.db import migrations, models
def update_snapshottag_ids(apps, schema_editor):
Snapshot = apps.get_model("core", "Snapshot")
SnapshotTag = apps.get_model("core", "SnapshotTag")
num_total = SnapshotTag.objects.all().count()
print(f' Updating {num_total} SnapshotTag.snapshot_id values in place... (may take an hour or longer for large collections...)')
for idx, snapshottag in enumerate(SnapshotTag.objects.all().only('snapshot_old_id').iterator()):
assert snapshottag.snapshot_old_id
snapshot = Snapshot.objects.get(old_id=snapshottag.snapshot_old_id)
snapshottag.snapshot_id = snapshot.id
snapshottag.save(update_fields=["snapshot_id"])
assert str(snapshottag.snapshot_id) == str(snapshot.id)
if idx % 100 == 0:
print(f'Migrated {idx}/{num_total} SnapshotTag objects...')
class Migration(migrations.Migration):
dependencies = [
('core', '0050_alter_snapshottag_snapshot_old'),
]
operations = [
migrations.AddField(
model_name='snapshottag',
name='snapshot',
field=models.ForeignKey(blank=True, db_column='snapshot_id', null=True, on_delete=django.db.models.deletion.CASCADE, to='core.snapshot'),
),
migrations.AlterField(
model_name='snapshottag',
name='snapshot_old',
field=models.ForeignKey(db_column='snapshot_old_id', on_delete=django.db.models.deletion.CASCADE, related_name='snapshottag_old_set', to='core.snapshot', to_field='old_id'),
),
migrations.RunPython(update_snapshottag_ids, reverse_code=migrations.RunPython.noop),
]

27
archivebox/core/migrations/0052_alter_snapshottag_unique_together_and_more.py Normal file
View file

@ -0,0 +1,27 @@
# Generated by Django 5.0.6 on 2024-08-20 02:37
import django.db.models.deletion
from django.db import migrations, models
class Migration(migrations.Migration):
dependencies = [
('core', '0051_snapshottag_snapshot_alter_snapshottag_snapshot_old'),
]
operations = [
migrations.AlterUniqueTogether(
name='snapshottag',
unique_together=set(),
),
migrations.AlterField(
model_name='snapshottag',
name='snapshot',
field=models.ForeignKey(db_column='snapshot_id', on_delete=django.db.models.deletion.CASCADE, to='core.snapshot'),
),
migrations.AlterUniqueTogether(
name='snapshottag',
unique_together={('snapshot', 'tag')},
),
]

17
archivebox/core/migrations/0053_remove_snapshottag_snapshot_old.py Normal file
View file

@ -0,0 +1,17 @@
# Generated by Django 5.0.6 on 2024-08-20 02:38
from django.db import migrations
class Migration(migrations.Migration):
dependencies = [
('core', '0052_alter_snapshottag_unique_together_and_more'),
]
operations = [
migrations.RemoveField(
model_name='snapshottag',
name='snapshot_old',
),
]

18
archivebox/core/migrations/0054_alter_snapshot_timestamp.py Normal file
View file

@ -0,0 +1,18 @@
# Generated by Django 5.0.6 on 2024-08-20 02:40
from django.db import migrations, models
class Migration(migrations.Migration):
dependencies = [
('core', '0053_remove_snapshottag_snapshot_old'),
]
operations = [
migrations.AlterField(
model_name='snapshot',
name='timestamp',
field=models.CharField(db_index=True, editable=False, max_length=32, unique=True),
),
]

15
archivebox/core/models.py
View file

@ -121,8 +121,8 @@ class Tag(ABIDModel):
class SnapshotTag(models.Model):
id = models.AutoField(primary_key=True)
snapshot = models.OneToOneField('Snapshot', on_delete=models.CASCADE, to_field='old_id')
tag = models.ForeignKey(Tag, on_delete=models.CASCADE, to_field='id')
snapshot = models.ForeignKey('Snapshot', db_column='snapshot_id', on_delete=models.CASCADE, to_field='id')
tag = models.ForeignKey(Tag, db_column='tag_id', on_delete=models.CASCADE, to_field='id')
class Meta:
db_table = 'core_snapshot_tags'
@ -135,12 +135,12 @@ class Snapshot(ABIDModel):
abid_subtype_src = '"01"'
abid_rand_src = 'self.old_id'
old_id = models.UUIDField(primary_key=True, default=uuid.uuid4, editable=False) # legacy pk
id = models.UUIDField(default=uuid.uuid4, editable=True, unique=True)
old_id = models.UUIDField(default=uuid.uuid4, editable=False, unique=True) # legacy pk
id = models.UUIDField(primary_key=True, default=uuid.uuid4, editable=True, unique=True)
abid = ABIDField(prefix=abid_prefix)
url = models.URLField(unique=True, db_index=True)
timestamp = models.CharField(max_length=32, unique=True, db_index=True)
timestamp = models.CharField(max_length=32, unique=True, db_index=True, editable=False)
title = models.CharField(max_length=512, null=True, blank=True, db_index=True)
@ -365,6 +365,7 @@ class ArchiveResult(ABIDModel):
EXTRACTOR_CHOICES = EXTRACTOR_CHOICES
old_id = models.BigIntegerField(default=rand_int_id, serialize=False, verbose_name='Old ID')
id = models.UUIDField(primary_key=True, default=uuid.uuid4, editable=True, unique=True, verbose_name='ID')
abid = ABIDField(prefix=abid_prefix)
@ -392,6 +393,10 @@ class ArchiveResult(ABIDModel):
super().save(*args, **kwargs)
assert str(self.id) == str(self.abid.uuid)
@property
def uuid(self):
return self.id
@cached_property
def snapshot_dir(self):
return Path(self.snapshot.link_dir)